This career development award is being sought by a critical care physician nearing the completion of his clinical training who is fully committed to a career as an independent, academic investigator studying ways to measure and improve patient safety in the intensive care unit (ICU). The candidate is soon to become a tenure-track Assistant Professor at a large University with an exceptional track record of training health services researchers. He will devote 80% of his effort to developing his research program and 20% to caring for critically ill patients. A three-part career development plan will fill gaps in the candidate's training by focusing his scholarship on patient safety research, training him to use advanced statistical methods to define associations between safety measures and outcomes, and introducing him to ethnographic methods in preparation for future studies of the mechanisms underlying medical errors. Throughout this award, he will be mentored by senior experts in health services research, patient safety, biostatistics, and ethnography. He will therefore be well-positioned to apply, in the latter half of the award, for independent research support (e.g., R01's) for multidisciplinary studies of patient safety that stem naturally from his mentored research. The research portion of this career development award will proceed in two phases. During phase 1, the investigator will determine the convergent and content validity of several plausible measurements of a new construct called ICU capacity strain, defined as a systemic and time-varying influence on an ICU's ability to provide high-quality care for everyone who is or could become a patient in that ICU on a given day. By prospectively collecting data at 2 large ICUs over a 1-year period and conducting targeted surveys of critical care safety experts, the candidate will select the measure of ICU capacity strain with the optimal combination of validity and ease of measurement. This measure will then be used as the primary exposure variable in a series of analyses using an existing dataset including nearly 300,000 patients admitted to 102 ICUs across the U.S. from 2002 - 2008. These analyses will evaluate the influence of ICU capacity strain on several metrics of patient safety: 1) failure to deliver evidence-based prophylactic therapies, 2) failures to perform tests designed to diagnose adverse events, and 3) risks for in-hospital mortality following transitions of care. In evaluating each of these relationships, the candidate will seek to identify threshold levels of ICU capacity strain beyond which patient safety deteriorates, organizational characteristics of ICUs that enable them to safely accommodate increased capacity strain, and whether increasing strain preferentially exposes the elderly to medical errors. This research will therefore produce a valid measurement of a previously unexplored threat to patient safety, an evidence base to inform current triage and staffing decisions, and a conceptual framework that non-ICU investigators may use to evaluate threats posed by capacity strain within other clinical entities. The increasing numbers of critically ill patients engendered by the aging population threatens the safety of critical care for all. By establishing a measurement of the strains placed on ICUs at times of high demand for their services, and determining how, when, and where such strains lead to medical errors, this work will help to improve the safety of care delivered to the more than 10 million Americans admitted to ICUs each year.